ABSTRACT OF PARENT GRANT Though life expectancy in the US has steadily increased and many people live to age 90 and beyond, health and quality of life is extremely poor for most. While Alzheimer's disease and related dementias (ADRD) affect 15% of those age 65+, by age 90+, this number increases to a startling 40-50%. The oldest-old, people aged 90+, are the fastest growing segment of the elderly population in the US, currently comprising 4.7% and expected to increase to 11.5% by 2060. Yet there's an enormous dearth of information on the epidemiology of mild cognitive impairment (MCI) and ADRD in the oldest-old, particularly in non-whites and those from lower socioeconomic classes. This is highly problematic as the proportion of non-white minorities is rapidly increasing and by 2060 will represent 36% of the oldest-old. Dementia and MCI rates highly vary between ethnic groups at younger ages yet it is completely unknown if patterns are the same in the oldest-old. Brain imaging and neuropathology studies so far in oldest-old suggest that vascular pathologies, rather than AD, play a larger role in dementia, yet this hasn't been examined in non-Whites. Beyond the lack of demographic diversity in 90+ studies of oldest-old, there is a paucity of information on early and midlife risk and protective factors for ADRD and brain pathology in oldest-old. Lifecourse studies in the oldest-old are imperative; yet very costly and logistically challenging since studies encompassing multiple decades are needed. Thus, careful investigation of lifecourse health and protective mechanisms is strongly needed in this high risk population to tease apart which factors, at what point in time, may protect an individual. We propose an unprecedented epidemiologic study of MCI and Dementia in the oldest-old whom we have 30-50 years of prospectively collected life history and health data. Kaiser Permanente has an unparalleled cohort of almost 7121 individuals currently aged 90+ (36% Non-White) who participated in the Multiphasic Health Study (MHC) with baseline exams from 1964-1973, and follow-ups to 1991. These data, joined with granular electronic medical records from 1996?present provide an exceptional and comprehensive resource. Eight hundred MHC members aged 90+ (600 Non-White and 200 White) without dementia will enroll in a study of incident dementia and MCI. Structural MRI and Amyloid PET will be obtained on a random subsample of 200 individuals (150 Non White and 50 White) to characterize cerebral amyloid burden, vascular lesions, and atrophy. Brain donation for postmortem pathology will be sought from all 800 participants. Our overall objectives are to estimate incidence of dementia/MCI in a diverse cohort of oldest-old, identify midlife and late-life risk and protective factors, and understand the pattern of cerebral and brain pathologies in this diverse oldest-old population.